FAA expands AI advice to manage air traffic congestion across US airspace

The FAA is expanding SMART, an $875 million cloud-based AI system, from an initial deployment over Washington, DC, to all 22 air route traffic control centers in the United States. It analyzes weather, flight paths, traffic flow, and controller staffing to forecast congestion and recommend schedules, routes, and aircraft arrival order. The agency says that since its September launch in the National Capital Region, the system has helped reroute 180 flights, saving an average of 10 minutes and nearly 80 gallons of fuel per flight. The FAA hopes nationwide use will reduce delays and fuel consumption. SMART has already been extended to centers serving Atlanta, Jacksonville, and Miami; one cited example involved a cargo plane receiving a shorter, faster route to Dallas-Fort Worth.

The system offers advice rather than directly controlling aircraft: FAA specialists in Washington review its recommendations before relaying them, and local controllers may accept or reject them. The agency says it cannot replace controllers’ critical role. Airline leaders have endorsed the rollout, but the National Air Traffic Controllers Association, representing more than 15,000 workers, says it had no part in SMART’s design, testing, or implementation and cannot assess its effects on the workforce. Public evidence of its performance remains limited: the FAA has not identified the cargo flight in its example or explained how the system has affected passenger traffic. The expansion comes as the agency faces controller staffing shortages, making questions about human oversight and workplace impact especially relevant.